With advancements in machine learning (ML) disrupting the future of work, Amazon has not only been working with students to educate them about the latest trends in the field of ML, and apply them to solve real-world problems, but Amazon is also providing students unique opportunities to boost their ML skills.
Hackathons are a great way to achieve this since they simulate project-based learning environments to sharpen technical skills. Students have an opportunity to work in a collaborative environment on real-world problems, accruing the benefits of peer-based learning.
The launch of the Amazon’s first-ever, week-long Machine Learning (ML) Challenge began on July 30 on an enthusiastic note. The flagship competition in ML was open to all students enrolled in PhD/ M.E./M.Tech./ M.S./B.E./B.Tech. graduating in 2022 or 2023 from engineering campuses in India. The finale, which was streamed live on Twitch saw the top 10 teams presenting their solutions to Amazon scientists.
The objective was to spur interest in ML among the students, encourage them to take up careers in ML and pursue research, by providing a glimpse into how we leverage ML to solve business problems in the e-commerce domain at Amazon.
Building innovative solutions for real world problems
Speaking on the occasion, Rajeev Rastogi, VP of India Machine Learning at Amazon, said, “Amazon ML Challenge provided participants a unique opportunity to work on Amazon’s dataset and build innovative solutions for a real world problem. While designing the hackathon, the objective was to spur interest in ML among the students, encourage them to take up careers in ML and pursue research, by providing a glimpse into how we leverage ML to solve business problems in the e-commerce domain at Amazon. We are delighted by the overwhelming response from students across the country. A big congratulations to all the winners!”
Collaborative learning experience
Sharing that building a pipeline of students for Machine Learning roles across tech campuses in India is a "high priority" for Amazon, Suman Yadav, who heads the Amazon Student Programs team for APAC said, "After the successful completion of 3-day ‘ML Summer School’ in July, Amazon launched the ‘ML Challenge’- a competition open to all tech students that provided them with a unique opportunity to apply theoretical knowledge and build innovative solutions for a real world business problem. The event saw registration from 12K+ students and 7K+ students submitted their solutions. The experience shared by participants increased their exposure to peculiar ML problems. The collaborative learning experience among team members was encouraging to see and we would like to continue this initiative in the coming years.”
How the ML challenge worked
During the two stage competition, students were provided a unique opportunity to work on Amazon’s dataset to bring in fresh ideas and build innovative solutions for a real world problem statement. In Round 1 of the ML Hackathon, students collaborated in teams of three or four to work on the given problem statement, which was to classify products into Browse Nodes.
Dealing with so much real-world data was unusual for us...We learnt a lot of things, tried a lot of things. It will be a memory to cherish.
The top three teams at finale were declared winners. Each of the winning teams received pre-placement interviews (PPIs) for ML roles at Amazon along with cash prizes and certificates. The ‘no_bert_train_challenge’ team from IIT ISM Dhanbad came out on top as they effectively showcased the use of cutting-edge machine learning techniques to solve the challenging problem of predicting product categories. The ‘DeVaSh.Ai’ from IIT Guwahati and the ‘Patanjali Noodles’ team from IIT ISM Dhanbad bagged the second and third place, respectively.
A learning experience for students
As 7000+ students from 1000+ colleges participated in the competition, we caught up with a winning team member to ask him about his experience. He said, "It was a great learning experience for me. Dealing with so much real-world data was unusual for us but we eventually came up with a strategy to make use of the entire training data. We learnt a lot of things, tried a lot of things. It will be a memory to cherish.”
While one student said that the challenges posed by the raw dataset was what he really liked, another said that "working on a real life problem helped him learn new concepts and handle large datasets.”
Earlier, in June 2021, Amazon India had also launched an ML Summer School to provide an integrated learning experience for students, and help them gain applied Machine Learning (ML) skills. ML Summer School is open to engineering students in their pre-final and final year Bachelors, Masters, or PhD studies at select 20 campuses. The curriculum for the three-day course covered fundamentals of machine learning in areas such as Deep Learning and Probabilistic Graphical Models to solve various industry-specific business problems.
These events will continue to be hosted annually by Amazon, and will thereby play a crucial role in upskilling the ML student community in India.